4.4 Discussion
4.4.1 The relative weight of linguistic typological distance and psychotypological distance
The results of OT, which aimed to examine beginning L3 learners’ productive skills, showed a clear-cut pattern. The result indicated that L3 beginning learners produced significantly more Japanese sentences in the Korean word order than in the Chinese word order in productive skills. Considering that beginning learners rely on the existing language(s) to facilitate the processing of the target language (Flege &
Davidian, 1984; Macken & Ferguson, 1981, among others), the result has confirmed that Korean plays a dominant role than Chinese in terms of productive skills at the L3 initial state. Therefore, it can be concluded that a linguistically-typologically closer language plays a dominant role than a psychotypologically closer language at the L3 initial state in productive skills.
In addition, the same pattern was exhibited across all the four types of sentence complexity. This suggests that beginning L3 learners successfully processed a morphosyntactic item that is much beyond their current proficiency within the time limit of 15 seconds. In fact, although the SOV/SVO sentence and the double-object sentences in OT belong to elementary level (i.e., N5 in JLPT), quotational sentences consisting of a quotational clause belong to lower-intermediate level (i.e., N4 in JLPT). Considering the fact that all the participants’ Japanese proficiency is a beginning level, there might be a possibility that their limited cognitive capacity could not allow them to produce complete quotational sentences. Contrary to this assumption, although some learners did not produce a complete sentence, overall,
is much beyond their current level at the L3 initial state. In this study, Korean promoted morphosyntactic transfer to L3 more dominantly than Mandarin Chinese in processing quotational sentences as well as the canonical SOV/SVO sentence and the double-object sentence.
Then, a question arises: what generated the clear pattern of Korean dominancy over Chinese in transferring to L3? It would probably be due to L3 learners’
possessing explicit awareness on morphosyntactic closeness between Korean and Japanese. Interestingly, the survey of psychotypology revealed that over 95% of the participants have already had conscious awareness that Korean word order is close to Japanese one. In other words, L3 learners, irrespective of Taiwanese or Korean learners, have already had the explicit knowledge on linguistic typological closeness between Korean and Japanese despite the beginning stage of learning Japanese as L3.
Despite their beginning stage of learning Japanese, the reason they have an explicit awareness would be that bilinguals who have attained advanced L2 proficiency have the relatively high-level metalinguistic awareness. Even so, it is misleading to claim that Korean knowledge was fully transferred to Japanese without any Chinese transfer.
In fact, Chinese transfer was also observed, although not so much. In this sense, it can be claimed that knowledge on Mandarin Chinese played a certain amount of role in misleading the L3 learners in processing the target Japanese sentences. However, the statistical analysis has verified that the explicit idea on word order of Korean played a major role in guiding L3 beginning learners to make use of the existing Korean linguistic knowledge to orally produce Japanese sentences.
In summary, the study clearly showed that a linguistically-typologically language played a major role in acquiring productive skills at the L3 initial state rather than a psychotypologically-close language. The outcome would probably stem from
L3 learners’ explicit awareness on linguistic similarity between the existing language and the target language. L3 learners could successfully transfer a linguistically-typologically close language due probably to high metalinguistic awareness.
4.4.2 The relative weight of linguistic typological distance and psychotypological distance in receptive skills
As found in OT, the statistical analysis on GJT showed that the effect of linguistic typology outweighed that of psychotypology. The within-subject effects clearly illustrated that, in both of the two experimental groups, Korean played a crucial role in processing Japanese as L3 than Mandarin Chinese irrespective of sentence complexity. Also, the result was consistent between the two experimental groups as the between-subjects effects showed no significant difference. This substantiates that a typologically-closer language influenced acquisition of receptive knowledge of the target L3 language more significantly than a psychotypologically-closer regardless of four types of sentence complexity. To account for this result, probably learners’ explicit awareness in (morpho)syntactic similarities between Korean and Japanese might play a role as the survey revealed that 95% of the participants have already been metalinguistically aware of the linguistic typological closeness.
Then, to what degree did the linguistically-typologically-close language assist in facilitating L3 acquisition of receptive skills? Interestingly, beginning L3 learners have attained the same level of accuracy as the native Japanese control group with the
speaker controls. The outcomes of a post hoc analysis showed that no significant difference was observed between L1 Korean/L2 Chinese learners and Japanese natives regardless of sentence complexity, and neither between L1 Chinese/L2 Korean learners and Japanese natives, although negligible exceptions were found. The statistical results collectively showed that L3 learners have achieved native-like receptive knowledge in L3 at their elementary level.
To sum up, in receptive skills as well as productive skills, the role of a linguistic typological distance was dominant over that of a psychotypological distance at the L3 initial state. The results offered the insight on revealing the relative weight of actual typological distance and psychotypological distance at the L3 initial state.